| Purpose: By observing features of ischemic stroke patients and analyzing risk factors of the disease recurrence based on classification trees model,this research aimed at providing further proofs to the prevention and treatment of ischemic stroke and giving a methodological reference to the usage of clinical database and data mining.Method: Select 1661 cases from January 1st 2015 to December 31 th 2016 in the Chronic Disease Prevention and Treatment Information Coordinate Platform of the Affiliated Hospital to Changchun University of Chinese Medicine.Divide the cases into 2 groups according to medical history.There are 771 cases of new ischemic stroke and 890 cases of recurrent ones.Analyze features of 2 groups by using descriptive statistical tests,logistic regression,classification trees model and discuss risk factors of ischemic stroke recurrence.Result: Single factor Logistic Regression: Compared the 2 groups,they have statistic difference at gender;age;history of hypertension,diabetes,CHD;vegetable diet;drink often;WHR and TCM syndromes.(p<0.05 or p<0.01)Multiple factor Logistic Regression: Features that differ the 2 groups are primary school and education degree below;WHR>0.9;medicine history of hypertension,diabetes,CHD;family history of CHD and syndromes of FengTanZuLuo,QiXuXueYu,TanReFuShi,TanMengQingQiao and TanReNeiBi.Classification Trees: Variables affect the recurrence are patriarchal CHD history;family care;history of hypertension,diabetes,CHD;WHR;gender;age and income.Variables used in optimal tree are family care;patriarchal CHD history;WHR;history of hypertension,diabetes and CHD.Conclusion: Compared with the first-episode ischemic stroke patients,recurrent patients have difference at education status,WHR,medical history and TCM syndromes.Family care,patriarchal coronary heart disease history,WHR,history of hypertension,diabetes,coronary heart disease are the most high risk factors of ischemic stroke recurrence. |
PDF Full Text Request